Further develop LEMI pole-zero fitting ipython notebook

[kkappler]

The notebook path is:
aurora/tutorials/pole_zero_fitting/pole_zero_fitting/lemi_pole_zero_fitting_example.ipynb

What I would like to do is:

1. Show multiple fits for {one pole} , {two pole one zero}, {three pole two zero}, etc using different initial conditions.
2. Review why convergence happens at different misfits for different initial conditions (could be a tolerance that needs to be set in scipy.optimize)
3. Add Jacobian and Hessian and run with 'Newton-CG' and some other optimizers
4. Review the physical significance of the PZ solutions that fit the model better but do not honour conjugate pairs of roots - what errors might we encounter by using these? Are they a practical solution